Freezing coils causing $20K floods!

[jcdenver]

Hoping I can get some suggestions and maybe some a more technical explaination on this...

We manage a 273 unit apartment complex in Denver, CO. Units are equipped with split package A/C's with hydronic coils for heat.

(these are vertically mounted units, return air enters from below, pulled up through evap coil across fan pushed through hydronic coil and out to apartment supply ducts. Hydronic coil is fed from water heater, via circ pump when heat is called for, but is static during a/c mode)

Every time the temps start to get low at night or during day in spring/fall we have issues with the evaporator freezing over creating an ice block, which the fan then sits there spinning freezing cold air around and through the hydronic coil ultimately freezing the water and popping the coil and causing $10-20K worth of flood damage by time we get the call from the resident downstairs that waters pouring out of every light fixture, crack and crevice in his apartment.

What we think is happening is the outside temps are low enough to cause the refrigerant to stack up in the condenser, starving the evap and causing it to freeze (Does this sound right? What is the limit for OA over the condenser?)


You'd think they wouldn't run a/c when outside temps get low but its a large building with alot of heat build up and many of these folks like to live like penguins.


We've installed a freeze kit on the suction line inside the air handler that is set to cut power to the condenser if temps hit 38 degrees. But a few of these have failed, have seen them sitting there in a block of ice with condenser still running.

Short of pulling the power to the disconnect everytime the temp drops outside to 273 units and listening to constant complaining are there any other options. Someone mention some time of unit that mounts on the condenser and can cycle the fan off and on to prevent the refrigerant from stacking in the condenser...?

[beenthere]

There are low ambient kits available to install on the condensers. They work, if its not caused by other problems.


At what outdoor temp does this start to happen.
70?
65?
60?

Are you sure you have enough air flow over the indoor coils. Low air flow can cause them to freeze up at 70° oudoor temp, but not at 75.

Some of those cheap freeze stats you put on the vapor/suction line won't hold up to constant cycling like you are probably getting.

[allan38]

It's called a low ambient kit. Companies that have computer server rooms have them on their AC system. Same situation, need AC when it's cold outside.

That isn't a cheap fix but cheaper than a $20K flood.

[cmajerus]

besides low ambient kits, those hydronic coils should have been mounted before the cooling coil. Most likely not going to happen now, but just pointing it out.

[Airmechanical]

Hoping I can get some suggestions Every time the temps start to get low at night or during day in spring/fall we have issues with the evaporator freezing over creating an ice block, which the fan then sits there spinning freezing cold air around and through the hydronic coil ultimately freezing the water and popping the coil and causing $10-20K worth of flood damage


my suggestion to you is to get a qualified HVAC contractor (with good insurance policy)

then put the problem in there hands, get them to write a proposal for fixing your problem

make sure the proposal is specific by stating their solution to fixing your specific problem!



.

[bmathews]

besides low ambient kits, those hydronic coils should have been mounted before the cooling coil. Most likely not going to happen now, but just pointing it out.

We've put a couple of Goodman hydronic units in. The heater coil sits on top, the evap. down low with the blower in the middle. It's the way their made. You might try a tstat with an outside temp. lockout on it.

[RoBoTeq]

besides low ambient kits, those hydronic coils should have been mounted before the cooling coil. Most likely not going to happen now, but just pointing it out.
Those units are often installed with heat pumps. If you put the hydronic coil before the DX coil, you would need to install a relay to shut off the heat pump whenever the hydronic coil was called to operate, thus loosing the ability to use the heat pump to it's fullest potential as well as reducing to overall heat available.

[RoBoTeq]

jcdenver, what are the make and models of the indoor and outdoor units?

Have you actually seen that the DX coils are frozen when the hydronic coils fail or are you going by what someone is telling you must be the case?

Where on the hydronic coils are the leaks? Do they have multiple splits in the end loops of the coil? Are they leaking from the coil/fin pack? Has the pump cracked?

[BamaCool]

[quote=beenthere;2011930]
Are you sure you have enough air flow over the indoor coils. Low air flow can cause them to freeze up at 70° oudoor temp, but not at 75.

I know you know what you're talking about, but you're saying low airflow can't cause a freezing problem if the OAT is above 75 ???:confused::confused:

[skippedover]

There are 3 things to look at here.

#1 Refrigerant charge - a low charge will freeze the AC coil and if the hydro coil is in close proximity, could freeze the water in it. Proper charging is the basic solution. An additional safety would be a discharge air temperature switch to turn off the AC condenser when the supply air temp dips below 50^F.
#2 AC system is called to run at a low outdoor ambient, which results in the same symptom as a low refrigerant charge. Solution is a low ambient kit or put on the supply discharge air switch and tell the HO's to open their windows when it's 60-degrees outdoors.
#3 Low airflow. Whether it's due to dirty air filters, dirty blower, plugged coils, newspaper on the coil, feather duster caught in the blower, whatever the reason, airflow sufficiently low will cause the AC coil to freeze. Again, a DAT switch could save the day.

Personally, I'd hire a highly qualified HVAC company to determine the best fix for your situation and hire them to do the job. I'd also, if it were my responsibility, put a DAT switch on every unit to cover all the bases if something goes wrong somewhere else in the system. Any system can leak refrigerant over a period of time for many different reasons. Why take a chance? DAT switch will save the day.

[RoBoTeq]

Here is the problem I have with the claim that a hydronic coil, located above a DX coil, with a blower assembly in between them; can freeze.

The only way the DX coil can freeze is if it drops to 32- degrees (I put the minus sign there because the temperature must drop to 32 degrees minus the latent heat required to solidify the condensate) at which point the DX coil is going to start freezing the condensate into ice. At the most, you will get 32 degree air pulled across the DX coil until the DX coil freezes completely, at which time no air will be blown across the hydronic coil. Since hot air rises and cold air falls, the colder air from the frozen DX coil is not going to rise past the blower assembly to freeze up the hydronic coil. Even if there is a cooler affect from the DX block of ice being the bottom of the cooler and the hydronic coil being 16" above it with no top, how is the hydronic coil going to drop below freezing?

[skippedover]

I agree with you 100% Robo. The hydro coil should never freeze as a result of the DX coil. But if there's fresh air coming in from outdoors in sub-freezing temps, that could freez a hydro coil that's not kept hot. Otherwise I was just trying to provide a solution to whatever the unseen problem is. Put a DAT sensor in the plenum and let it go. That should cover all the bases, no?

[referteacher]

Am I missing something, here :confused:????

Every hydronic system I've been involved with has GLYCOL in the loop.:rolleyes:

Even a minimal amount would keep it from freezing, per RoBoTeq's explanation, above.

[skippedover]

Am I missing something, here :confused:????

Every hydronic system I've been involved with has GLYCOL in the loop.:rolleyes:

Even a minimal amount would keep it from freezing, per RoBoTeq's explanation, above.

You're not missing anything up to this point but if the hyrdoair coil wasn't sized for glycol, then it's likely to be undersized if glycol is added. One has to desigin for glycol in a system due to the loss of heat transfer. Adding it as an afterthought can bring significant headaches.

[RoBoTeq]

The systems being referred to are mostly installed to be fed by water heaters. Since the water is being fed from the same source as the water heater is, glycol is not allowed into the system. The water that goes through the hydronic system is returned to the water heater.

I have documentation, as well as actual coils, on coils that have burst when the cooling system was not operational and the temps never reached freezing. I have documentation on coils that have burst on units that were located in interior rooms that have never had freezing conditions. I believe there is something else occuring rather then the alleged frozen coil excuse.

This is why I am curious as to what brand and model of hydronic coil has failed and where on the coil the failure is.

[just_opinion]

I don't know how these A/C units can spit out 32-28 degrees temp to freeze the steam coil down to the 28 - 29 degrees for the water to expand then burst the steam coil.

I may under stand that the steam coil got cold and water condensed, then all of the sudden the massive amount of steam introduce that causes rapid expansion of the water, which rupture the steam coil. I may understand that.

So I don't think freeze A/C coils cause the freeze then cause these $20K problem.

[RoBoTeq]

I don't know how these A/C units can spit out 32-28 degrees temp to freeze the steam coil down to the 28 - 29 degrees for the water to expand then burst the steam coil.

I may under stand that the steam coil got cold and water condensed, then all of the sudden the massive amount of steam introduce that causes rapid expansion of the water, which rupture the steam coil. I may understand that.

So I don't think freeze A/C coils cause the freeze then cause these $20K problem.
They are hot water coils, not steam coils. Regardless, I believe you are correct that the freezing to the point of bursting the copper, especially in the manner I have documentation of them bursting out of several end loops, cannot occur.

[Shophound]

If the coil is not bursting due to freezing temperatures, is it possible that, although the water does not freeze, the cold atmosphere created by the frozen DX coil contracts the hydronic coil metal/copper to the extent a weak joint or connection gives way? Seems counter-intuitive, I know, being that joints or connections would seem to give more under pressure than contraction.

Where exactly are the hydronic coils failing? Is it in the same location, more or less, each time?

How the coils are failing will tell a great deal about possible causes for the failure. Freeze failures generally result in the copper being literally split open, vs. pinhole leaks or joint separation.

Also, if the hydronic coil was freezing, it would not flood the apartment until both the DX coil and the hydronic coil were thawed out. If you walk into an apartment with a burst hydronic coil, spewing water, that water is likely hot if it's from the hot water system with no form of isolation from the hot water boiler (solenoid or zone valve).

Do these failures ever occur when the system is in heating mode?

What about installing zone valves on the hydronic coils so they are not under pressure when not in use?

[Shophound]

One other thought regarding cause of DX coils freezing...has the total external static pressure been checked since this system has two coils on it vs. one? Such a check should be done with the evaporator coil cold and wet, system running in cooling mode.

Apartment tenants (and many times maintenance staff) are notoriously bad about changing filters. Two coils in a dirty airstream mean two coils that can collect dirt. The first one way more than the other, but whatever amount of dirty air makes it past the first coil due to bypass factor will foul the second. Tenants should also never shut off any supply vents in the apartment. These systems may have, brand new, run just barely inside the TESP window they are designed to deliver proper airflow. Let the filter and coils foul up with dirt and it's just downhill from there.

[RoBoTeq]

In cases that I have investigated, which result in millions of dollars in damages, the coils all have multiple split open end loops on both sides of the coil. Usually 11-15 end loops will be split open as if blown out from the inside. These have occured in cooling and in heating modes. When they occur in cooling mode, the DX coil is blamed. When they occur in heating mode, freezing temperatures are claimed. I have quite a bit of evidence that disputes any of these claims.

[Shophound]

In cases that I have investigated, which result in millions of dollars in damages, the coils all have multiple split open end loops on both sides of the coil. Usually 11-15 end loops will be split open as if blown out from the inside. These have occured in cooling and in heating modes. When they occur in cooling mode, the DX coil is blamed. When they occur in heating mode, freezing temperatures are claimed. I have quite a bit of evidence that disputes any of these claims.

If the hydronic coil is failing but not due to being frozen, what would cause catastrophic failure of several end loops at once? I'm not disputing your findings, I'm curious to learn what is causing the failures outside of a freeze-up situation. Are we speaking of formicary corrosion on the inside of the coil? Or is it an interaction with the tube sheet and end loops (expansion/contraction)?

I'll put this out there, since I have experience dealing with dry pipe fire sprinkler manifolds that have sprung leaks due to microbiologically induced corrosion, aka MIC. This is where a microbe present in city water literally eats away at the pipe's surface. The failures occured where standing water was left in an otherwise dry pipe pressurized with air. The failures also occurred under the standing water, not in the air sections of the pipe. We have used dehydration and standing nitrogen pressure to prevent a reoccurrence of failures due to MIC, along with a biocide. So far, so good.

It could be MIC is present in the hydronic coil, being that it is using city water from the hot water heater. Still baffles me as to multiple failures on one coil, if it did not freeze. But...could also be u-bends were weak in several areas, but not under pressure, and when the circ pump was started, the sudden hydraulic shock and increase of pressure ruptured the u-bends.

[RoBoTeq]

If the hydronic coil is failing but not due to being frozen, what would cause catastrophic failure of several end loops at once? I'm not disputing your findings, I'm curious to learn what is causing the failures outside of a freeze-up situation. Are we speaking of formicary corrosion on the inside of the coil? Or is it an interaction with the tube sheet and end loops (expansion/contraction)?

I'll put this out there, since I have experience dealing with dry pipe fire sprinkler manifolds that have sprung leaks due to microbiologically induced corrosion, aka MIC. This is where a microbe present in city water literally eats away at the pipe's surface. The failures occured where standing water was left in an otherwise dry pipe pressurized with air. The failures also occurred under the standing water, not in the air sections of the pipe. We have used dehydration and standing nitrogen pressure to prevent a reoccurrence of failures due to MIC, along with a biocide. So far, so good.

It could be MIC is present in the hydronic coil, being that it is using city water from the hot water heater. Still baffles me as to multiple failures on one coil, if it did not freeze. But...could also be u-bends were weak in several areas, but not under pressure, and when the circ pump was started, the sudden hydraulic shock and increase of pressure ruptured the u-bends.
I have been trying to figure this out for more then five years. I believe it has to do with static electrical discharge, a collegue believes it has to do with water pressure along with internal wear on the inside of the end tubes due to friction of the water flow. Since the blowouts are all in the very end of the end loop, actually in the furthest point of the loops themselves, my collegue believes that internal wear along with the fact that this is the thinnest portion of the end loop due to the bending process, is causing multiple bursts in end loops similtaneaosly. I disagree. I feel that only an electrical discharge will be so precisely timed as to blow out so many end loops at the same time.

The freezing scenario would also explain multiple points of damage because it could occur over a longer period of time while the water is solidified in the coil. Aside from the issue of these happening during times where freezing conditions just don't make sense, I have done my own freeze experimentation on these coils. I have filled one of these coils with water, pressurized it to 6 psi and let sit in my reach in freezer over night. The result was multiple coil bursts. The difference is that in my frozen coil, not one end loop burst. All of the damage was to the straight sections of tubing in the coil fin pack.

I'll see if I can dig up some of my photos of some of these coils. In the meantime, I really wish the OP would come back and tell us particulars on his coil leaks.

[Shophound]

Static electrical discharge is an interesting angle...would there not be traces of arc flash at the failure locations if it was due to static electricity? Every time I have seen high voltage interact with metal/copper/steel, there are traces of arc burns/charring. For multiple failures on the same coil I would think the voltage needed to blow several return bends simultaneously would be pretty high...such as a lightning strike or direct short to ground of a high voltage electrical load or wiring.

We know thinner copper walls are used on DX coils to increase heat transfer...is there a similar push for hydronic coils? If so, I can see internal erosion/corrosion causing one u-bend to fail, but multiple? Simultaneously? That's what is puzzling.

[BaldLoonie]

The water coils are fed by the apartment's water heater. Pretty hard to use glycol :eek:

Simple strap on temp control set about 40 should keep the evap from freezing on top of the fan cycle kit since they run the units in cold weather.

[RoBoTeq]

Here are some photos of different coils I have seen burst open;

[just_opinion]

Robo Teq,

I have the same problem with commercial building too. They are brand new coils and they burst just like that. They coil got burst is the one at the highest point of the loop (boiler is in the lowest floor). Even on the old coil got bursted just like that.

Once again, you and I agreed that the A/C cannot get that hydronic coil froze up and bursted. Look and see how far apart these A/C coil and water coil are apart. And there is NO WAY this A/C unit spits out freezing temperature that long to freeze the coil over to burst it.

Is this unit located at the highest point of the loop??

[RoBoTeq]

There are three different units represented in the pics I posted. All of them have the hot water feeding the coils coming from water heaters on the same floor in another closet location.

Notice that there is plastic plumbing piping and BX cable feeds to the air handling units. I think that some of these systems have an inadequate grounding and the static buildup in the hydronic coils is aplified by the water and discharges through the coil tubes to the meal unit casing, which has a better grounding due to the DX coil being copper piped to the outdoor unit.

A sudden burst of low amp, high voltage electrical charge is the only thing I can think of that could blow out this many spots at the same time. The theory of the water pressure doing it just doesn't make sense because once the first opening occured, the pressure would drop, preventing subsequent loops from bursting. Besides; how much pressure are you really going to get from a water system compared to the pressure of a refrigerant system?

I also have photos of what appears to be burn spots on the pump impellars right at the shaft location. I have been told those are rust spots, but that is not conclusive with my testing of these spots.

As shown, the end of the pump has also cracked completely. This also happened with the coil that I froze in my freezer, so I am assuming this could occur under different circumstances. Not one end loop showed any signs of breach with my frozen test coil. All breaches were in the straight sections of tubing within the fin pack.

[Frostie]

The bursts in the coils in your pictures look just like frozen pipe bursts. Wouldn't the evaporator being frozen be proof enough that the coils froze up and burst?

[RoBoTeq]

The bursts in the coils in your pictures look just like frozen pipe bursts. Wouldn't the evaporator being frozen be proof enough that the coils froze up and burst?
In the dozen or so incidences I have investigated, a frozen DX coil was never actually found, only theorized. Also, the same exact thing has happened in winter months on inside located air handlers as well on air handlers that were not exposed to sub-freezing temperature days.

If it were not for so many contradictions, I would have agreed with the frozen coil scenario and not have tested a coil myseld by freezing it. My test coil did not burst in any of the end loops.

[Shophound]

I would still expect to see arc burns, as if an arc welder had been there, for the cause to be from an electrical surge. And...I would expect the openings to be irregular shaped holes vs. the clean splits seen in the u-bends.

Going with your info regarding your recorded cases, I will maintain the line of thought that freezing did not burst these coils. My inclination would then be toward hydraulic pressure. If these systems did not have an expansion tank installed on the loop, the only thing left to happen for water inside a pipe to do when heated, with no air cushion for a buffer, is to find the weakest points in the loop and burst them. Water, being a liquid, is not compressable, so if the rate of expansion and heating exceeds the ability of the containing vessel to withstand it, the vessel bursts.

Yes, it would seem logical that the first bend that gives would instantly relieve the pressure, so no other u -bends should burst. However, the amount of pressure needed to actually burst the u -bend would no doubt be substantial, and felt at many u-bends simultaneously. Having about the same ability to resist the pressure, the u-bends burst simultaneously when the pressure becomes excessive. I would also hold that the rate of expansion might be dynamic (that is, continuing to increase) when the first u-bend bursts, which would not reduce the rate sufficiently to prevent others from bursting. It is only when multiple bursts have occurred that the remaining u-bends are spared.

I have seen arc shorts burn through EMT and copper pipe. Never have seen same cause multiple elongated slits in copper pipe. If there is no expansion tank on these hot water loops, I would heavily lean my thoughts that direction.

[RoBoTeq]

I would still expect to see arc burns, as if an arc welder had been there, for the cause to be from an electrical surge. And...I would expect the openings to be irregular shaped holes vs. the clean splits seen in the u-bends.

Going with your info regarding your recorded cases, I will maintain the line of thought that freezing did not burst these coils. My inclination would then be toward hydraulic pressure. If these systems did not have an expansion tank installed on the loop, the only thing left to happen for water inside a pipe to do when heated, with no air cushion for a buffer, is to find the weakest points in the loop and burst them. Water, being a liquid, is not compressable, so if the rate of expansion and heating exceeds the ability of the containing vessel to withstand it, the vessel bursts.

Yes, it would seem logical that the first bend that gives would instantly relieve the pressure, so no other u -bends should burst. However, the amount of pressure needed to actually burst the u -bend would no doubt be substantial, and felt at many u-bends simultaneously. Having about the same ability to resist the pressure, the u-bends burst simultaneously when the pressure becomes excessive. I would also hold that the rate of expansion might be dynamic (that is, continuing to increase) when the first u-bend bursts, which would not reduce the rate sufficiently to prevent others from bursting. It is only when multiple bursts have occurred that the remaining u-bends are spared.

I have seen arc shorts burn through EMT and copper pipe. Never have seen same cause multiple elongated slits in copper pipe. If there is no expansion tank on these hot water loops, I would heavily lean my thoughts that direction.
This is more the train of thought of one of my counterparts who has also been involved with these coils bursting. To add to his theory, he feels that the minerals in the water would have further weakened the already thinned end loops making them even more succeptable to failure.

Either way, the freezing theory is the least probable. The reason for my leaning more toward a burst of static discharge energy is because these coils have failed in the cooling season as well as the heating season and always on systems where plastic water piping breaks the grounding through the water pipe system.

[BURL-REF]

About 30 years ago there was a supermarket that had water cooled compressors. They went through many halsted michell type condensors. They had found they were having electrolosis thru the phone companies ground was the problem..

[temco]

We had situation with large Trane air handler with hot water coil and DX coil added later. Twice the hot water coil bend had burst on same coil at same area. Trane engineers said it was because of DX coil freezing water. Coils were close together. I saw the second coil, it was burst just like Robos pictures except at the area of burst it was also swelled up. I thought maybe the somehow the condenser was coming on during unoccupied time when airflow was off. Not the case.
Talked with engineer with Hydronic experience. He thought maybe water hammering was causing problem along with weak, thin, velocity worn area of coil. This was the last, very furthest point of four other Trane coils installed at same time on the top floor of two story building. Burst was on far side of coil. Out of four this was the only one with problem (twice). It turns out that public work dept. had worked on main water lines at same time building flooded. When they open the main water lines they had previous problems with breaking components from pressure and water hammering. Im not sure how air or pressure could make it thru requlators, make up water lines, etc and damage coil. But I guess it doesnt take a huge spike or jolt in pressure to send a pocket of air down the line at high velocity. Im still not sure if this is what happened. But we installed hammer arrestors at direction of engineer at before the last coil about five years ago and they havnt blown yet.
I still have no real idea what happened but its interesting to see Robos pictures and hear other similar problems happened at the end of the line or top of the piping loop.

[rich pickering]

In the dozen or so incidences I have investigated, a frozen DX coil was never actually found, only theorized. Also, the same exact thing has happened in winter months on inside located air handlers as well on air handlers that were not exposed to sub-freezing temperature days.

If it were not for so many contradictions, I would have agreed with the frozen coil scenario and not have tested a coil myseld by freezing it. My test coil did not burst in any of the end loops.

When coils freeze, the fan is usually running. The coils freeze in the middle and the ice plug expands, compressing the water, which bursts the ends of the coils which are also the weakest part of the system. Most of the time, I find ice still in the coil, but nothing in the return bends.

[RoBoTeq]

We had situation with large Trane air handler with hot water coil and DX coil added later. Twice the hot water coil bend had burst on same coil at same area. Trane engineers said it was because of DX coil freezing water. Coils were close together. I saw the second coil, it was burst just like Robos pictures except at the area of burst it was also swelled up. I thought maybe the somehow the condenser was coming on during unoccupied time when airflow was off. Not the case.
Talked with engineer with Hydronic experience. He thought maybe water hammering was causing problem along with weak, thin, velocity worn area of coil. This was the last, very furthest point of four other Trane coils installed at same time on the top floor of two story building. Burst was on far side of coil. Out of four this was the only one with problem (twice). It turns out that public work dept. had worked on main water lines at same time building flooded. When they open the main water lines they had previous problems with breaking components from pressure and water hammering. Im not sure how air or pressure could make it thru requlators, make up water lines, etc and damage coil. But I guess it doesnt take a huge spike or jolt in pressure to send a pocket of air down the line at high velocity. Im still not sure if this is what happened. But we installed hammer arrestors at direction of engineer at before the last coil about five years ago and they havnt blown yet.
I still have no real idea what happened but its interesting to see Robos pictures and hear other similar problems happened at the end of the line or top of the piping loop.
These coils are all on individual systems fed by an individual water heater. On all coils, end loops at both ends of the coil burst. As you mentioned, if there was a freezing condition; where are the swells in the copper? There are none on the coils that have burst while in operation but on the test coil that I froze, the straight tubing within the fin pack that burst had swelled significantly.

[RoBoTeq]

When coils freeze, the fan is usually running. The coils freeze in the middle and the ice plug expands, compressing the water, which bursts the ends of the coils which are also the weakest part of the system. Most of the time, I find ice still in the coil, but nothing in the return bends.
But there have been no absolutes and mostly the contrary to any freezing conditions. If the outdoor ambient is not below freezing, are you saying that a DX coil 18" below a hot water coil with a blower between them can discharge a low enough temperature to freeze that hot water coil? How? By the time the DX coil reaches below freezing temperatures, it is itself freezing up and blocking the air.

Also, no one that I know of has ever been able to recreate this frozen coil scenario.